* docs: revert to labels={"foo.bar": "baz"} style
Back in #24074 I thought it was necessary to wrap labels in a list to
support quoted keys in hcl2. This... doesn't appear to be true at all?
The simpler `labels={...}` syntax appears to work just fine.
I updated the docs and a test (and modernized it a bit). I also switched
some other examples to the `labels = {}` format from the old `labels{}`
format.
* copywronged
* fmtd
The go-metrics library retains Prometheus metrics in memory until expiration,
but the expiration logic requires that the metrics are being regularly
scraped. If you don't have a Prometheus server scraping, this leads to
ever-increasing memory usage. In particular, high volume dispatch workloads emit
a large set of label values and if these are not eventually aged out the bulk of
Nomad server memory can end up consumed by metrics.
* fix(doc): fix links for task driver plugins
host URL was wrong, changed from develoepr to developer
* Update stateful-workloads.mdx
Fix link for Nomad event stream page
In https://github.com/hashicorp/nomad/issues/15459 we've had a bit of
back-and-forth as a result of applying Nomad environment variables where they
typically should not be used. Clarify that the env vars are for the CLI and
mostly not for the agent. Also move the `NOMAD_CLI_SHOW_HINTS` description into
the correct section.
The docs for the `template` block accurately describe the template configuration
default function denylist in the body but the default parameters are missing
values. The equivalent docs in the `client` configuration are missing
`executeTemplate` as well.
* Add -log-file-export and -log-lookback commands to add historical log to
debug capture
* use monitor.PrepFile() helper for other historical log tests
* Add MonitorExport command and handlers
* Implement autocomplete
* Require nomad in serviceName
* Fix race in StreamReader.Read
* Add and use framer.Flush() to coordinate function exit
* Add LogFile to client/Server config and read NomadLogPath in rpcHandler instead of HTTPServer
* Parameterize StreamFixed stream size
Affinities and contraints use similar feasibility checking logic to determine if
a given node matches (although affinities don't support all the same
operators). Most operators don't allow `value` to be unset. Update the docs to
reflect this.
Fixes: https://github.com/hashicorp/nomad/issues/24983
During the big docs rearchitecture, we split up the task driver pages into
separate job declaration and driver configuration pages. The link for the
`raw_exec` driver to the configuration page is a self-reference.
The documentation for CSI and DHV has a list of the available access modes, but
doesn't explain what they mean in terms of what jobs can request, the scheduler
behavior, or the CSI plugin behavior. Expand on the information available in the
CSI specification and provide a description of DHV's behavior as well.
Ref: https://github.com/container-storage-interface/spec/blob/master/spec.md#createvolume
The current autoscaler docs implies that it has minimal or non-working support
for Nomad namespaces. Whereas in fact the namespace support works fine but just
doesn't allow configuring multiple namespaces without using a wildcard (for
now). Make this more clear and fix the reference to the configuration "below",
which is no longer on that same page.
Ref: https://github.com/hashicorp/nomad-autoscaler/issues/65
The `DesiredUpdates` struct that we send to the Read Eval API doesn't include
information about disconnect/reconnect and rescheduling. Annotate the
`DesiredUpdates` with this data, and adjust the `eval status` command to display
only those fields that have non-zero values in order to make the output width
manageable.
Ref: https://hashicorp.atlassian.net/browse/NMD-815
System and sysbatch jobs don't support the reschedule block, because we'd always
replace allocations back onto the same node. The job validation for system jobs
asserts that the user hasn't set a `reschedule` block so that users aren't
submitting jobs expecting it to be supported. But this validation was missing
for sysbatch jobs.
Validate that sysbatch jobs don't have a reschedule block.
If you delete a CSI volume, the volume cannot be currently claimed by an
allocation or in the process of being unpublished. This is documented in the CLI
but not the API. Also, the documentation incorrectly says that the `volume
delete` command silently returns without error if the volume doesn't exist, but
that's incorrect.
Fixes: https://github.com/hashicorp/nomad/issues/24756
When we originally implemented CSI, Nomad did not support the `CreateVolume`
workflow, so the volume name field was just a display name. The `CreateVolume`
CSI RPC requires that the volume name be unique. In retrospect, Nomad should
probably have mapped the namespace + ID to the volume name field, but because we
didn't the name field must be unique per storage provider. In future work we
should try to figure out a way to unwind that decision but in the meantime let's
make that requirement clear in the documentation.
Ref: https://gitlab.com/rocketduck/csi-plugin-nfs/-/issues/21
